New-to-Country Consumers Represent Trillions in Economic Value. But First, Their Identities Must Be Validated.

The potential impact of immigration on US economic growth is astounding. According to research conducted by FWD.us, high rates of legal immigration between now and 2050 will result in approximately $50 trillion in GDP. With low rates of immigration, the US will likely abdicate its title as global economic leader by 2030, and will have about $37 trillion in GDP by 2050. 

Irrespective of how government policy on immigration shapes up over the coming decades, it’s clear that those who are new-to-country—which currently account for almost 14% of the population—bring with them significant buying power as participants in the US economy. Enabling their transformation as consumers, however, will require that financial institutions, retailers, and other organizations have an accurate way to validate and onboard these individuals so they can participate in the US economy.

This important segment of the population is clearly valuable from an economic standpoint, but also presents identity verification challenges. Their financial and credit histories may be thin, or even non-existent. Many lack the standard identification documentation that most enterprises require to onboard new applicants. How can banks and retailers validate these users without the risk of introducing fraud into their systems?

These challenges can be overcome with the right approach to identity verification. Let’s consider some of the current issues surrounding this population, and see how organizations can use identify verification to overcome them. 

New-to-country populations have both a data-coverage and an entity-resolution challenge that legacy providers simply can’t address

Additionally, identity verification and fraud mitigation technologies play an important role in helping financial services organizations reach certain populations in an increasingly digital world. Unfortunately, the rapid digitization of identity verification practices has led to unintended bias in the decisioning of many legacy solutions. 

Unfair bias occurs when an organization uniformly applies a practice (e.g., identity risk assessment) that has a discriminatory effect on certain segments of the population to all applicants. An example of this could be a person of a certain age or demographic unnecessarily being subjected to increased friction during onboarding. Though this friction may not be intentional, it could be the result of hidden biases in machine learning algorithms, in the data sample that is gathered and used for modeling, as well as in the entity resolution techniques deployed. 

This same type of bias occurs in new-to-country populations as well. This group typically does not have a record in the credit reporting ecosystem, so identifying these individuals requires legacy vendors to use various step-up methods and other time-consuming actions to check additional data sources. The result is that these consumers often experience substantially higher friction in the application and onboarding processes, reduced opportunities in credit availability, and generally are denied access to the consumer services available to those with established US credit histories. 

New-to-country populations are more likely to use different surname and first name characteristics, as well as hyphenated names and apostrophes. These standard international naming conventions often don’t work well with the decades-old legacy forms of entity resolution in place here in the US. 

The disparity in the performance of legacy providers can be explained through their overreliance on static edit distances, soundex phonetic encoding, and rigid lexical rules that have long been favored in entity resolution problem solving. The problem with these types of entity resolution is that it often leads to negative bias regarding Latin, Middle Eastern, and Asian names, which may entail hyphenation or short string length. 

It is important to note that technology access, while often thought to be a factor, is not itself a driver of inequality, rather it is the onboarding speed (or lack thereof) and friction caused by antiquated CIP programs and biased fraud mitigation strategies that create unequal access to financial services.

Organizations that are serious about delivering inclusive onboarding must operate with a solution that moves beyond the limited scope of the analytical capabilities of legacy identity verification solutions. 

Socure identifies and serves more new-to-country consumers better than any other identity provider

Rather than relying on the accepted best practices of the past, Socure has combined the maximum amount of data to cover all ages, races, demographics, and genders with best-in-class entity resolution, clustering, and machine learning (ML) technology to resolve and correlate to any given identity with the highest degree of accuracy and coverage.

Well-governed models are key to achieving highly accurate identity outcomes. Socure has made substantial data science investments to drive precision, accuracy, and greater pass rates for CIP programs, and to optimize the coverage of identity information that informs our proprietary highly accurate machine-learning fraud models for these new-to-country populations.

We have tested our models against the results of legacy identification providers and can prove substantial lift in accurate identification for new-to-country populations, including pass rate lifts of 36% for Hispanic consumers, and 46% for Asian consumers.

Organizations that can foster greater inclusion for credit invisibles not only promote positive social change, but tap into and build a market of long-term, loyal customers. Failure to do so means they will not effectively reach important growing segments of the population, which will result in missed opportunities for revenue, customer growth, and development of a positive reputation within the communities they serve.

Read our Digital Identity Fairness & Inclusion Report to learn more.